Commit graph

1218 commits

Author SHA1 Message Date
Mateo Wang
4eb13a0b2a
Merge pull request #41843 from BerriAI/litellm_lit8064_unpin_derived_pricing
fix(proxy): unpin cost-map pricing copied into model_info and report pricing overrides
2026-09-18 22:58:04 -07:00
Mateo Wang
078a60478f
Merge pull request #41485 from BerriAI/litellm_jwt_token_exchange_grant
feat(proxy): add RFC 8693 token exchange for IdP JWTs on the gateway token endpoint
2026-09-18 21:27:20 -07:00
ryan
f3bbeed82f feat(proxy): let team admins manage projects via team_admin_editable_team_fields
Adds a projects entry to the team_admin_editable_team_fields setting. When set, team admins (legacy admins list or members_with_roles role admin) can call /project/new and /project/update for the teams they administer. The two routes join self_managed_routes so the endpoint check runs instead of the route gate's blanket 401. /project/delete stays proxy admin only

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-19 01:02:50 +00:00
mateo-berri
50629ff5ca Merge branch 'main' of https://github.com/BerriAI/litellm into litellm_jwt_token_exchange_grant
# Conflicts:
#	tests/test_litellm/proxy/auth/test_auth_checks.py
2026-09-18 17:30:38 -07:00
ryan-crabbe-berri
1073b9eff7
Merge pull request #41354 from BerriAI/litellm_lit_3269_project_spend_tracking
fix(proxy): track project spend and enforce project budgets additively
2026-09-18 17:17:28 -07:00
mateo-berri
f143a46cdf chore: merge main into litellm_lit8064_unpin_derived_pricing 2026-09-18 16:11:37 -07:00
Yassin Kortam
87694c26ef
Merge pull request #41324 from BerriAI/litellm_daily_global_spend_table
feat(proxy): add LiteLLM_DailyGlobalSpend key-free rollup for the usage dashboard
2026-09-18 14:53:08 -07:00
ryan-crabbe-berri
a5f6ce7bb1 Merge remote-tracking branch 'origin/main' into litellm_lit_3269_project_spend_tracking
# Conflicts:
#	tests/test_litellm/proxy/db/test_db_spend_update_writer.py
2026-09-18 14:22:07 -07:00
ryan
61d4c5b9b5 fix(proxy): skip members already on the team before resolving a per-member budget
A mixed /team/member_add list that names an existing member used to run
add_new_member for them, which created or cloned a budget that the empty
upsert update branch never linked to their membership row. Filter the
requested members against the freshly locked roster first so budgets and
membership rows are only written for members who are actually new

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 13:54:17 -07:00
ryan
de4b520153 fix(proxy): track team member spend when the member has no budget
add_new_member only wrote a LiteLLM_TeamMembership row when a budget id
resolved, and the spend writer used update_many so a missing row failed
silently. Members without a budget therefore never accrued per-member spend.

The membership row is now always upserted (budget_id NULL when no budget
applies) and the spend write is an upsert so members added before this fix
start accruing on their next request.

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 13:54:16 -07:00
yassin
737c6e1b2d Merge remote-tracking branch 'origin/main' into litellm_mcp_admin_terminate_sessions_revoke_credentials
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

# Conflicts:
#	tests/test_litellm/proxy/config_resolvers/test_settings_store.py
2026-09-18 20:32:31 +00:00
yassin
6a6ae2d064 Merge remote-tracking branch 'origin/main' into litellm_daily_global_spend_table 2026-09-18 20:20:17 +00:00
Yassin Kortam
2e46b10320
Merge pull request #41293 from BerriAI/litellm_usage_key_free_aggregate_split
perf(proxy): split aggregated usage query into key-free rollups and bounded top-N keys
2026-09-18 13:16:49 -07:00
yassin
1feffc3635 Merge remote-tracking branch 'origin/main' into litellm_mcp_admin_terminate_sessions_revoke_credentials 2026-09-18 20:16:22 +00:00
ryan-crabbe-berri
6e3b6d6d03
Merge pull request #41830 from BerriAI/litellm_scim_multivalued_optional_value
fix(scim): accept entitlements and roles entries without a value on SCIM user PUT
2026-09-18 13:13:51 -07:00
yassin
b92820dcfe Merge branch 'litellm_usage_key_free_aggregate_split' into litellm_daily_global_spend_table 2026-09-18 20:06:47 +00:00
yassin
191ca14872 Merge remote-tracking branch 'origin/main' into litellm_usage_key_free_aggregate_split 2026-09-18 20:05:54 +00:00
ryan-crabbe-berri
6b54238083
Merge pull request #40807 from BerriAI/litellm_service_account_team_key_mgmt
feat(keys): let team service account keys use key management endpoints for their own team
2026-09-18 12:58:41 -07:00
yassin
307df09792 test(mcp): assert the self-revoke response instead of echoing the delete mock
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 18:51:43 +00:00
mateo-berri
42541a9233 fix(proxy): drop echoed cost-map pricing on a row's next save and build /model/info pricing stamps without mutation 2026-09-18 11:42:07 -07:00
yassin
048aaad627 Merge remote-tracking branch 'origin/main' into litellm_mcp_admin_terminate_sessions_revoke_credentials 2026-09-18 18:39:16 +00:00
Yassin Kortam
0594dd7caf
Merge pull request #41539 from BerriAI/litellm_vault_login_secret_namespace
feat(vault): add separate login and secret namespaces for HashiCorp Vault
2026-09-18 11:38:49 -07:00
Yassin Kortam
84ae0805ba
Merge pull request #41620 from BerriAI/litellm_team_member_temp_budget_increase
feat(proxy): temporary budget increase for team members
2026-09-18 11:38:34 -07:00
Yassin Kortam
f6d9b2552f
Merge pull request #41692 from BerriAI/litellm_mcp_gateway_sessions_by_client_user
feat(mcp): show live gateway sessions by AI client and user
2026-09-18 11:37:16 -07:00
Yassin Kortam
ca79c393d5
Merge pull request #41636 from BerriAI/litellm_per_key_end_user_default_budget
feat(proxy): per-key default budget for dynamically created customers
2026-09-18 11:36:59 -07:00
kerry-berri
fd58c31cc7
Merge pull request #41832 from BerriAI/litellm_lit_8111_cache_read_missing_rate
fix(cost): bill cache-read tokens at the input rate when the map has no cache-read rate
2026-09-18 10:56:25 -07:00
mateo-berri
b03957ba9c fix(proxy): unpin cost-map pricing copied into model_info and report pricing overrides
A model_info blob that carries key next to pricing fields is a copy of a /model/info response (only litellm.get_model_info emits key), so those pricing fields are dropped when the row is loaded from the DB and on every Reload Price Data, and the deployment follows the current cost map again. Prices typed into litellm_params, or into model_info without key, stay as they are.

/model/info, /v1/model/info and /v2/model/info now report model_info.pricing_overrides, the pricing fields the deployment sets itself, and the Admin UI model page says whether a price follows the cost map or overrides it.
2026-09-18 10:49:36 -07:00
mateo-berri
b523ed9a2d merge: origin/main into litellm_jwt_token_exchange_grant
Some checks failed
LiteLLM Rust / rust-lint (push) Has been cancelled
LiteLLM Rust / rust-test (push) Has been cancelled
LiteLLM Rust / rust-wheel (push) Has been cancelled
2026-09-18 10:35:53 -07:00
yuneng-jiang
c4ab1d98e9
Merge pull request #41779 from BerriAI/litellm_settings_store_precedence
refactor(proxy): make the config file win over the database
2026-09-18 09:52:09 -07:00
kerry
af61d9247c test(cost_tracking): expect cache reads without a map rate to estimate at the input rate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 16:33:17 +00:00
ryan
817cdcef64 test(scim): drop redundant docstring from the valueless entitlements PUT test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 16:00:54 +00:00
ryan
b6410d563b fix(scim): accept entitlements and roles entries without a value on SCIM user PUT
SCIMMultiValuedAttribute required value, so a PUT /scim/v2/Users/{id} that
carried an IdP-specific entitlements entry such as {"groups": [...]} failed
body validation with 422 and the suspend (active: false) never reached
update_user. value is now optional and unknown members are kept, so the
suspend is applied, keys are blocked, and the entries are stored as sent

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 15:46:33 +00:00
ryan-crabbe-berri
4f70b88a1f
Merge pull request #41707 from BerriAI/litellm_jwt_mapping_cache_evict_on_bulk_key_delete
fix(proxy): evict jwt key mapping cache on user, team, org, and bulk key deletion
2026-09-18 08:41:31 -07:00
yassin
f631301cfa Merge branch 'litellm_usage_key_free_aggregate_split' into litellm_daily_global_spend_table 2026-09-18 09:29:47 +00:00
shivam
a052da6974 feat(keys): let team service account keys use key management endpoints for their own team
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 09:05:10 +00:00
yassin
3d2913307a Merge remote-tracking branch 'origin/main' into litellm_usage_key_free_aggregate_split 2026-09-18 08:59:24 +00:00
yassin
bc636bc5ff Merge remote-tracking branch 'origin/main' into litellm_vault_login_secret_namespace 2026-09-18 08:57:50 +00:00
Yuneng Jiang
afa4a6fe78
refactor(proxy): make the config file win over the database
The precedence used to vary per key: some keys let a stored row win, some
let the file win, some merged the two. That meant an operator could not
answer "which value is live?" without knowing the key.

Now file presence decides ownership. A key the config file declares is
config-owned, whatever the database holds, and a key the file omits falls
back to the stored row. KeyRule no longer carries a RuleKind, only which
row the stored value lives in.

Writes to a config-owned key are refused at the two surfaces that reach
the database instead of being stored and silently ignored: save_config
and /config/field/update both 400 naming the key and the config file path.

Both read endpoints now report source and editable off the same
SettingsStore, so /config/field/info and /config/list can no longer
disagree inside one process.

Replaces the 786-case checked-in JSON fixture with cases generated from
the rule table, so the matrix tests no longer assert that resolve() agrees
with a snapshot of resolve().

BREAKING CHANGE: a dashboard or /config/field/update write to a setting
the config file declares now returns 400 instead of being stored. Remove
the key from the config file to let the database own it.
2026-09-18 01:28:21 -07:00
Yuneng Jiang
d1cd869012
refactor(proxy): resolve config and DB settings precedence in one SettingsStore 2026-09-17 23:36:27 -07:00
yassin
e6349d00f3 Merge remote-tracking branch 'origin/main' into litellm_mcp_admin_terminate_sessions_revoke_credentials 2026-09-18 02:10:58 +00:00
yassin
fd834f6f8b fix(mcp): broadcast BYOK and OAuth credential eviction to peer workers and expire admin session tombstones
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 02:10:50 +00:00
yassin
ea37596b88 feat(mcp): let proxy admins force-close live MCP sessions and revoke stored user credentials
Adds an admin-only DELETE /v1/mcp/sessions that terminates stateful MCP gateway sessions on the current worker by session id prefix and/or by the LiteLLM user that opened them, tombstones the terminated ids so a client reusing one gets 404 instead of a silently recreated stateless session, and lets PROXY_ADMIN name a user_id on the BYOK and OAuth credential delete routes. Full and view-only admins can list every user's stored credential metadata for a server (never the secret). The dashboard gains Disconnect controls on the Live Connections tab and a User Credentials tab with Revoke controls, both hidden from read-only admins.

Resolves LIT-8001

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 01:01:26 +00:00
mateo-berri
788e6eb737 merge: origin/main into litellm_jwt_token_exchange_grant 2026-09-17 17:59:23 -07:00
kerry-berri
f51f01fb54
Merge pull request #41443 from BerriAI/litellm_remove_brittle_price_pinning_tests
test: delete unit-test assertions that pin cost-map prices, limits and deprecation dates
2026-09-17 17:52:42 -07:00
ryan-crabbe-berri
85fe646776
Merge pull request #41694 from BerriAI/litellm_rename_model_sync_allowlists
fix(proxy): propagate db model renames to key, team, org, project and user model allowlists
2026-09-17 17:42:06 -07:00
kerry
8504c51f6c Revert "test: keep prompt cache prediction logic tests and drop only their price pins"
This reverts commit 810df257d8.
2026-09-18 00:40:38 +00:00
kerry
810df257d8 test: keep prompt cache prediction logic tests and drop only their price pins
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 00:39:34 +00:00
kerry
84d4d17bf8 Merge remote-tracking branch 'origin/main' into litellm_remove_brittle_price_pinning_tests
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

# Conflicts:
#	tests/test_litellm/llms/bedrock/chat/test_converse_transformation.py
2026-09-18 00:29:52 +00:00
kerry
c4620170ca test: delete assertions that pin vendor cost map facts
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 00:28:49 +00:00
ryan
696587c4ab fix(ui): persist disabling cache control injection points on model update
Turning Cache Control off on the model edit screen omitted the field from the
PATCH body, which the backend reads as leave unchanged, so the stored
cache_control_injection_points list survived the save. The dashboard now sends
an explicit null when a stored list is being disabled, and update_db_model
clears that field on null the same way it already clears the mirrored pricing
fields. Omitted keys keep the stored value

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-18 00:20:22 +00:00